Date: Tue, 13 Jan 2009 18:15:12 -0600
From: Steve Bergman <sbergman27@...il.com>
To: john-users@...ts.openwall.com
Subject: Re: Clarification desired on modifying incremental
 mode to handle 9 chars

On Sun, 2009-01-11 at 07:33 +0300, Solar Designer wrote:

> Thank you for posting this info - I imagine that some list members may
> find it useful.
> 
Things slowed down rather quickly from there.

> Here's an approach you may try.  With the following in your john.conf,
> you may use the bundled all.chr file to have JtR try 9 character
> candidate passwords:

All in all, after more consideration, I decided that it really was more
sensible to focus the processing power on the length <= 8 space.  What I
did end up doing, for comparison, after letting the Q6600 run for a
couple of days was to start the same thing again, but on an 8 core Xeon
E5320 (1.86GHz) box, and using the "Parallel" filter in the default
john.conf (suitably edited for 8 cores). I note that after a day and a
half of running it is allocating less processor time to the shorter
password lengths and more to lengths 7 and 8 than with the previous
method which divided the work up by password length. I find that better
suits my tastes.  I also tried the bindshell.net patched 1.7.2, which
worked well enough, but the lack of a restore capability is a show
stopper. At least, I think it lacks that capability. I believe I saw a
post of yours which mentioned that being a difficult problem, and I
didn't find any obvious way to do it. (My attempts to subscribe to the
john-mpi mailing list have been, thus far, unsuccessful, and I haven't
found an archive.)  The Parallel filter is obviously a bit of a hack,
but it seems to work pretty well for these relatively slow salted MD5
hashes, as is mentioned on the wiki.

-Steve  


-- 
To unsubscribe, e-mail john-users-unsubscribe@...ts.openwall.com and reply
to the automated confirmation request that will be sent to you.

Hosted by DataForce ISP - Powered by Openwall GNU/*/Linux